Continental-scale convection-permitting modeling of the current and future climate of North America

Climate Dynamics, Journal Year: 2016, Volume and Issue: 49(1-2), P. 71 - 95

Published: Aug. 29, 2016

Language: Английский

---

What Role Does Hydrological Science Play in the Age of Machine Learning?

Water Resources Research, Journal Year: 2020, Volume and Issue: 57(3)

Published: Nov. 14, 2020

Abstract This paper is derived from a keynote talk given at the Google's 2020 Flood Forecasting Meets Machine Learning Workshop. Recent experiments applying deep learning to rainfall‐runoff simulation indicate that there significantly more information in large‐scale hydrological data sets than hydrologists have been able translate into theory or models. While growing interest machine sciences community, many ways, our community still holds deeply subjective and nonevidence‐based preferences for models based on certain type of “process understanding” has historically not translated accurate theory, models, predictions. commentary call action hydrology focus developing quantitative understanding where when process valuable modeling discipline increasingly dominated by learning. We offer some potential perspectives preliminary examples about how this might be accomplished.

Language: Английский

---

Hydrologic refugia, plants, and climate change

Global Change Biology, Journal Year: 2017, Volume and Issue: 23(8), P. 2941 - 2961

Published: March 20, 2017

Abstract Climate, physical landscapes, and biota interact to generate heterogeneous hydrologic conditions in space over time, which are reflected spatial patterns of species distributions. As these distributions respond rapid climate change, microrefugia may support local persistence the face deteriorating climatic suitability. Recent focus on temperature as a determinant insufficiently accounts for importance processes changing water availability with climate. Where scarcity is major limitation now or under future climates, likely prove essential persistence, particularly sessile plants. Zones high relative – mesic microenvironments generated by wide array processes, be loosely coupled therefore buffered from change. Here, we review mechanisms that their robustness We argue will act species‐specific refugia only if nature space/time variability compatible ecological requirements target species. illustrate this argument case studies drawn California oak woodland ecosystems. posit identification could form cornerstone climate‐cognizant conservation strategies, but would require improved understanding change effects key including frequently cryptic such groundwater flow.

Language: Английский

---

Challenges in modeling and predicting floods and droughts: A review

Wiley Interdisciplinary Reviews Water, Journal Year: 2021, Volume and Issue: 8(3)

Published: March 11, 2021

Abstract Predictions of floods, droughts, and fast drought‐flood transitions are required at different time scales to develop management strategies targeted minimizing negative societal economic impacts. Forecasts daily seasonal scale vital for early warning, estimation event frequency hydraulic design, long‐term projections developing adaptation future conditions. All three types predictions—forecasts, estimates, projections—typically treat droughts floods independently, even though both extremes can be studied using related approaches have similar challenges. In this review, we (a) identify challenges common drought flood prediction their joint assessment (b) discuss tractable tackle these We group into four interrelated categories: data, process understanding, modeling prediction, human–water interactions. Data‐related include data availability definition. Process‐related the multivariate spatial characteristics extremes, non‐stationarities, changes in extremes. Modeling arise analysis, stochastic, hydrological, earth system, modeling. Challenges with respect interactions lie establishing links impacts, representing interactions, science communication. potential ways tackling including exploiting new sources, studying a framework, influences compounding drivers, continuous stochastic models or non‐stationary models, obtaining stakeholder feedback. Tackling one several will improve predictions help minimize impacts extreme events. This article is categorized under: Science Water >

Language: Английский

---

How the performance of hydrological models relates to credibility of projections under climate change

Hydrological Sciences Journal, Journal Year: 2018, Volume and Issue: 63(5), P. 696 - 720

Published: March 22, 2018

Two approaches can be distinguished in studies of climate change impacts on water resources when accounting for issues related to impact model performance: (1) using a multi-model ensemble disregarding performance, and (2) models after their evaluation considering performance. We discuss the implications both terms credibility simulated hydrological indicators adaptation. For that, we confirm hypothesis that good performance historical period increases confidence projected under change, decreases uncertainty projections models. Based this, find second approach more trustworthy recommend it assessment, especially if results are intended support adaptation strategies. Guidelines global- basin-scale period, as well criteria rejection from an outlier, also suggested.

Language: Английский

---

Nonstationary weather and water extremes: a review of methods for their detection, attribution, and management

Hydrology and earth system sciences, Journal Year: 2021, Volume and Issue: 25(7), P. 3897 - 3935

Published: July 7, 2021

Abstract. Hydroclimatic extremes such as intense rainfall, floods, droughts, heatwaves, and wind or storms have devastating effects each year. One of the key challenges for society is understanding how these are evolving likely to unfold beyond their historical distributions under influence multiple drivers changes in climate, land cover, other human factors. Methods analysing hydroclimatic advanced considerably recent decades. Here we provide a review drivers, metrics, methods detection, attribution, management, projection nonstationary extremes. We discuss issues uncertainty associated with approaches (e.g. arising from insufficient record length, spurious nonstationarities, incomplete representation sources modelling frameworks), examine empirical simulation-based frameworks analysis extremes, identify gaps future research.

Language: Английский

---

What Role Does Hydrological Science Play in the Age of Machine Learning?

EarthArXiv (California Digital Library), Journal Year: 2020, Volume and Issue: unknown

Published: Feb. 11, 2020

We suggest that there is a potential danger to the hydrological sciences community in not recognizing how transformative machine learning will be for future of modeling. Given recent success applied modeling problems, it unclear what role theory might future. central challenge hydrology right now should clearly delineate where and when adds value prediction systems. Lessons learned from history motivate several clear next steps toward integrating into workflows.

Language: Английский

---

The ClimEx Project: A 50-Member Ensemble of Climate Change Projections at 12-km Resolution over Europe and Northeastern North America with the Canadian Regional Climate Model (CRCM5)

Journal of Applied Meteorology and Climatology, Journal Year: 2019, Volume and Issue: 58(4), P. 663 - 693

Published: Jan. 30, 2019

Abstract The Canadian Regional Climate Model (CRCM5) Large Ensemble (CRCM5-LE) consists of a dynamically downscaled version the CanESM2 50-member initial-conditions ensemble (CanESM2-LE). downscaling was performed at 12-km resolution over two domains, Europe (EU) and northeastern North America (NNA), simulations extend from 1950 to 2099, following RCP8.5 scenario. In terms validation, warm biases are found EU NNA domains during summer, whereas winter cold appear NNA, respectively. For precipitation, generally wetter than observations but slight dry also occur in summer. change projections for 2080–99 (relative 2000–19) show temperature changes reaching 8°C summer some parts Europe, exceeding 12°C northern Québec winter. central will become much dryer (−2 mm day −1 ) (>1.2 ). Similar observed although drying is not as prominent. Projected interannual variability were investigated, showing increasing decreasing winter, Temperature increase by more 70% 80% northernmost part month May snow cover becomes subject high year-to-year future. Finally, CanESM2-LE CRCM5-LE compared with respect extreme evidence that higher allows realistic representation local extremes, especially coastal mountainous regions.

Language: Английский

---

ESD Reviews: Model dependence in multi-model climate ensembles: weighting, sub-selection and out-of-sample testing

Earth System Dynamics, Journal Year: 2019, Volume and Issue: 10(1), P. 91 - 105

Published: Feb. 13, 2019

Abstract. The rationale for using multi-model ensembles in climate change projections and impacts research is often based on the expectation that different models constitute independent estimates; therefore, a range of allows better characterisation uncertainties representation system than single model. However, it known groups share literature, ideas representations processes, parameterisations, evaluation data sets even sections model code. Thus, nominally might have similar biases because similarities way they represent subset or be near-duplicates others, weakening assumption estimates. If there are near-replicates some models, then treating all equally likely to bias inferences made these ensembles. challenge establish degree which this true any given application. While issue recognised by many community, quantifying accounting dependence anything other an ad-hoc challenging. Here we present synthesis disparate attempts define, quantify address common conceptual framework, provide guidance how users can test efficacy approaches move beyond weighted ensemble. In upcoming Coupled Model Intercomparison Project phase 6 (CMIP6), several new closely related existing anticipated, as well large from models. We argue quantitatively addition performance, thoroughly testing effectiveness approach used will key sound interpretation CMIP future scientific studies.

Language: Английский

---

Uncertainty estimation with deep learning for rainfall–runoff modeling

Hydrology and earth system sciences, Journal Year: 2022, Volume and Issue: 26(6), P. 1673 - 1693

Published: March 31, 2022

Abstract. Deep learning is becoming an increasingly important way to produce accurate hydrological predictions across a wide range of spatial and temporal scales. Uncertainty estimations are critical for actionable prediction, while standardized community benchmarks part model development research, similar tools benchmarking uncertainty estimation lacking. This contribution demonstrates that can be obtained with deep learning. We establish procedure present four baselines. Three baselines based on mixture density networks, one Monte Carlo dropout. The results indicate these approaches constitute strong baselines, especially the former ones. Additionally, we provide post hoc analysis put forward some qualitative understanding resulting models. extends notion performance shows learns nuanced behaviors account different situations.

Language: Английский

---